greedyPOP

PET Processing

**Content creator:** Katie Jobson

PET-only Centiloid Processing Pipeline

greedyPOP calculates Centiloid values from amyloid PET data without requiring a corresponding MRI scan. It is based on the rPOP (Robust PET-Only Processing) methodology and uses Greedy for image registration.

Description

greedyPOP performs the following steps:

  1. Origin correction (optional) - Centers the image origin if needed
  2. Skull stripping - Uses FreeSurfer’s SynthStrip for brain extraction
  3. Registration - Registers PET to tracer-specific templates using Greedy (affine + deformable)
  4. Quality control - Validates registration quality via Dice score comparison
  5. Smoothing - Estimates FWHM using AFNI and applies differential smoothing
  6. SUVR calculation - Computes standardized uptake value ratios
  7. Centiloid conversion - Converts SUVR to Centiloid scale using tracer-specific equations
  8. Visualization - Generates QC images and ITK-SNAP workspace

Supported Tracers

Tracer Trade Name
Florbetapir (FBP) Amyvid
Florbetaben (FBB) Neuraceq
Flutemetamol (FLUTE) Vizamyl

Reference Regions

Centiloid values are computed using multiple reference regions:

  • Whole Cerebellum (WhlCbl)
  • Cerebellar Gray Matter (CerebGry)
  • Pons
  • Whole Cerebellum + Brainstem (WhlCblBrnStm)

Installation

Docker (Recommended):

docker pull kjobson/greedypop:1.0.0

Build from Source:

git clone https://github.com/kjobson-neuro/greedyPOP.git
cd greedyPOP
docker build -t greedypop:latest .

Docker Usage 

docker run -v /path/to/data:/flywheel/v0/input \
           -v /path/to/output:/flywheel/v0/output \
           -v /path/to/work:/flywheel/v0/work \
           kjobson/greedypop:1.0.0 \
           -a /flywheel/v0/input/pet_scan.nii.gz \
           -r Florbetaben \
           -t Eight \
           -o Keep

Command Line Options

Flag Description Values
-a Path to PET data file (required) NIfTI file path
-r Tracer type (required) Florbetapir, Florbetaben, Flutemetamol
-t Target resolution Six (default), Eight, Ten
-o Origin setting Keep (default), Reset
-v Verbose mode (flag only)

Singularity / Apptainer

For HPC environments where Docker is not available:

# Pull and convert to SIF format
singularity pull greedypop_1.0.0.sif docker://kjobson/greedypop:1.0.0

# Run with Singularity
singularity run \
    --bind /path/to/data:/flywheel/v0/input \
    --bind /path/to/output:/flywheel/v0/output \
    --bind /path/to/work:/flywheel/v0/work \
    greedypop_1.0.0.sif \
    -a /flywheel/v0/input/pet_scan.nii.gz \
    -r Florbetaben \
    -t Eight \
    -o Keep

Inputs

  • PET data: NIfTI file (.nii or .nii.gz) or DICOM ZIP archive
    • If multi-volume, motion correction and averaging are applied automatically

Outputs

File Description
sw_pet.nii.gz Smoothed, warped PET image (template space)
suvr.nii.gz SUVR image in template space
suvr_native.nii.gz SUVR image in native space
greedyPOP_*.csv Results CSV with SUVR, Centiloid values, and FWHM estimates
greedyPOP.itksnap ITK-SNAP workspace for visualization
*.png QC visualization images

Software Dependencies

Other Resources

Citation

If you use greedyPOP in your research, please cite the original rPOP publication:

Iaccarino L, et al. rPOP: Robust PET-only processing of community acquired heterogeneous amyloid-PET data. NeuroImage. 2022;246:118775. doi: 10.1016/j.neuroimage.2021.118775

License

MIT License

Disclaimer: greedyPOP is distributed for academic/research purposes only, with NO WARRANTY. greedyPOP is not intended for any clinical or diagnostic purposes.